1. Field of the Invention
This invention relates to a method for making a semiconductor magnetoresistive device whose resistance varies with a changing applied magnetic field.
2. Description of Prior Art
A semiconductive magnetoresistor is a two-pole semiconductor device which may be substituted for a conventional resistor without any circuit alteration. The ohmic resistance of a magnetoresistor increases as the square of the magnetic field for weak magnetic fields and as a linear function of the magnetic field for larger magnetic fields. The resistance change is maximized in semiconductor materials in which the carrier mobility is high. Consequently, semiconductive magnetoresistive devices are normally made from a high mobility material such as indium antimonide.
The semiconductive magnetoresistor achieves its increased resistance in a magnetic field by an increase in the electrical current path length in a conducting specimen in a magnetic field. FIG. 1a shows the carrier path and equipotential lines in a conducting semiconductive magnetoresistor with metallic contacts in the absence of a magnetic field. FIG. 1b shows the electrical current path and equipotential lines with a magnetic field perpendicular to the plane of the drawing. The curvature of the electrical current path is caused by the Lorentz force of the magnetic field on the current carriers. The increased length in the electrical current path is the cause of increased resistance of the semiconductor magnetoresistor. The equipotential lines are rotated within the semiconductor body as charge builds up on the sides 1-2 and 3-4 of the specimen from the transverse component of the electrical current and generates an electrical field across the specimen. This electrical field is the source of the well known Hall voltage. The Hall voltage generated across the specimen opposes the further transverse flow of current carriers so that the current flow away from the metallic electrodes returns to a longitudinal flow along b-b' (FIG. 1a) thereby decreasing the current path length and thus the magnetoresistance. Near the electrodes both the current paths and the equipotential lines are affected by the much smaller carrier mobility and much higher carrier density of the metal electrodes. Because of the high carrier density and decreased carrier mobility in the metal electrodes, the average velocity of the carriers moving along the imposed electric field in the metal is much smaller than the average velocity of carriers in the semiconductor body where relatively very few highmobility carriers move at high velocities along the imposed electric field to carry the same amount of current as through the metal electrodes. Since the Lorentz force depends directly on the charged carrier velocity, it is negligible in the metal compared to the direct force of the imposed electric field. Consequently, the carriers move straight through the metal electrodes along longitudinal lines parallel to the sides 1-2 and 3-4, making the metallic electrodes equipotential lines.
In order to satisfy conditions of continuity between the metallic electrodes and the semiconductor body, both current and equipotential lines are rotated in the manner shown in FIG. 1b. This rotation increases the current path within about an electrode width in the semiconductor body and causes an increase in the resistance of the device. That portion of the semiconductor body more than about an electrode width away from the two electrodes does not increase the resistance of the device because the current path is not any longer in a magnetic field than in the absence of a magnetic field in this region, since in both cases the current is practically parallel to sides 1-2 and 3-4.
Consequently, to maximize the change in resistance of the device in a magnetic field, it is desirable to have a small spacing L between the metallic electrodes or their equivalent substitutes relative to their transverse dimensions W.
With a given size semiconductor body, this restriction suggest that a parallel series of metallic electrodes plates be placed at small interelectrode spacings in the semiconductor body in order to maximize the increase in the current path in a magnetic field.
One means of nearly achieving this arrangement in a practical magnetoresistance material such as InSb has been disclosed by Weiss (H. Weiss & N. Wilhelm, Z. Physik 176, 399 (1963). Weiss fabricated an InSb-NiSb eutectic crystal by directional solidification in which the NiSb is present as parallel needlelike inclusions 22 of a mean length of 50.mu. InSb crystal. These NiSb inclusions have a conductivity [7.times.10.sup.4 (.OMEGA.-cm.sup.-1)] two orders of magnitude greater than intrinsic or slightly n-doped antimonide [300 (.OMEGA.-cm).sup.-1 ].
A magnetoresistor can be made from this material by cutting out a section of the crystal and placing electrical contacts so that the parallel needle inclusions of NiSb are perpendicular to the current flow in the absence of a magnetic field. Without a magnetic field, the current flow is perpendicular to the applied voltage equipotential lines and flows straight through the specimen as shown in FIG. 2a. In a magnetic field perpendicular to the plane of the figure, the current no longer flows perpendicular to the applied voltage equipotential lines shown in FIG. 2a, but is deflected by the Lorentz force (charge X velocity X magnetic field) in direction more parallel to the needles as shown in FIG. 2b. Since the NiSb short-circuiting needles 22 are analogous to the metal electrodes discussed above, these needles 22 remain at a equipotential with the result that current injected into one portion of a needle 22 will flow towards the other parts of a needle 22 to maintain the equipotential along the needle 22. Thus the current flow will flow in a zig-zag path as shown in FIG. 2b. Because of the elongation of the current path in a magnetic field in FIG. 2b relative to the current path without a magnetic field in FIG. 2a, the device will have a higher two-pole resistance in a magnetic field than without a magnetic field.
Although the directional solidification of the InSb-NiSb eutectic produces a usable commercial magnetoresistor, factors associated with the use of this process limit the performance of the resulting magnetoresistor. These limitations are as follows:
(a) instead of the desired highly conducting planar morphology, the eutectic process produces the less desirable needle-like morphology for the highly conducting phase.
(b) the average needle length is only 50.mu. and thus does not extend across the whole width of the magnetoresistor, thereby reducing the performance of the magnetoresistor.
(c) the conductivity of the needle-like phase is the inherent conductivity of the nickel antimonide compound and cannot be increased to enhance the magnetoresistance effect.
(d) the carrier mobility in the resistor varies from values readily obtainable in bulk indium antimonide, thereby reducing the magnitude of the resistance change. This reduction in carrier mobility in the eutectic indium antimonide results from: (1) the stress and strains produced by the differential thermal contraction of the InSb and NiSb phases on cooling from the eutectic point, and (2) the inadvertent incorporation of trace impurities in the eutectic melt during the high-temperature solidification processing as well as the unavoidable incorporation of the solubility limit of nickel in the indium antimonide eutectic phase.
(e) the volume fraction of both semiconductor and conductor are fixed and predetermined when an eutectic is used os that this feature of an eutectic magnetoresistor cannot be optimized to give the desired ideal mixture of phases.
Therefore, it is an object of this invention to provide a new and improved method for making a semiconductor magnetoresistor which overcomes the limitations of the prior art devices.
Another object of this invention is to provide a new and improved method for making a semiconductor magnetoresistor with highly conducting planar regions in place of the current highly-conducting needle-like regions.
Another object of this invention is to provide a new and improved method for making a semiconductive magnetoresistor in which the dimensions of the highly conducting region are not limited and are such that the conduction region can extend across the entire specimen width.
Another object of this invention is to provide a new and improved method for making a semiconductive magnetoresistor in which the conductivity of the highly conducting phase can be varied to optimize device performance.
Another object of this invention is to provide a new and improved method for making a semiconductive magnetoresistor in which the carrier mobility in the semiconductive phase is as high as that available in large bulk crystals of the pure semiconductive phase.
Another object of this invention is to provide a new and improved method for making a semiconductive magnetoresistor in which the volume fraction of the conducting phase can be varied to optimize device performance.
Other objects of this invention will, in part, be obvious and will, in part, appear hereinafter.